[Ace-users] Re: [tao-users] DDS Question

[Adam Mitz]

Sowayan, Abdullah (N-DUA) wrote:
> Hi
>
> This isn't a question related to TAO. Though, I was hoping someone on
> this list might know the answer to this question.
>
> I started looking at some DDS products out there. DDS uses a subset of
> OMG IDL. OMG defined IDL to C++ mapping. Does DDS mandate using the OMG
> IDL to C++ mapping?
>   

The DDS spec doesn't mandate a given implementation language, but it 
would be reasonable to conclude that a DDS implementation in language A 
would use the OMG IDL mapping for language A.  For that matter, since 
the spec gets into this whole PIM/PSM business (which is mostly 
theoretical and is of little use practically when only one PSM is 
defined) you could have an implementation that uses an alternate PSM 
(not OMG IDL).

> The reason I ask is that I noticed that in that DDS products, their IDL
> compiler does NOT generate _var types, their infrastructure uses
> Foo::narrow instead of Foo::_narrow, the mapping of sequences is
> different than from OMG IDL to C++ mapping.
>
>   

Of course this isn't OpenDDS since we use TAO's IDL compiler!  :-)


> So I was wondering if that would be counted as a non-conformance point,
> or if OMG IDL to C++ mapping applies to CORBA only.
>   

It would seem to me that this would be an issue.  The whole idea of 
using OMG IDL within DDS is to allow interoperability of the code with 
existing CORBA solutions, and to leverage the knowledge and experience 
that developers have with the IDL mapping for their language of choice.

Thanks,
Adam